Video content analytics can be useful at airports... Ive recently been working on such a system at a busy US airport. In practicality, alot of the algorithms mentioned here are not useful. Walk into a busy airport any day of the week and you can count dozens of travellers leaving objects unattended. Also, if I were leaving a package with a bomb in it at an airport, I'd put it in a garbage can or underneath something. VCA would miss that. I've found the more useful algorithms are people/vehicles entering restricted zones and flow control.
Also, there is a misconception that VCA is most useful with real-time alerts. I work on very large-scale camera networks, and I've recently been convinced that the more valuable peice that VCA brings to the table is making sense of the large amounts of video that large scale networks produce. If you have a 1000 cameras each storing 120 days of video, if you are trying to search your video archives... thats 120,000 days of video! You need tools and meta-data around that video to make your archives more effective.
Finally... and Ive seen this time and time again... the video systems and VCA at airports are disjoint with the incident management systems they employ. If you cant respond effectively to the alarms in the video... what good is the video! You need integrated systems here.
Slashdot Mirror
http://truesentry.com/ and http://publiceye.silkblogs.com/ for more info
Video content analytics can be useful at airports... Ive recently been working on such a system at a busy US airport. In practicality, alot of the algorithms mentioned here are not useful. Walk into a busy airport any day of the week and you can count dozens of travellers leaving objects unattended. Also, if I were leaving a package with a bomb in it at an airport, I'd put it in a garbage can or underneath something. VCA would miss that. I've found the more useful algorithms are people/vehicles entering restricted zones and flow control.
Also, there is a misconception that VCA is most useful with real-time alerts. I work on very large-scale camera networks, and I've recently been convinced that the more valuable peice that VCA brings to the table is making sense of the large amounts of video that large scale networks produce. If you have a 1000 cameras each storing 120 days of video, if you are trying to search your video archives... thats 120,000 days of video! You need tools and meta-data around that video to make your archives more effective.
Finally... and Ive seen this time and time again... the video systems and VCA at airports are disjoint with the incident management systems they employ. If you cant respond effectively to the alarms in the video... what good is the video! You need integrated systems here.